Perform the reverse logistic regression estimation
revlogreg(lglk, N)A vector containing the reverse logistic regression estimates of the logarithm of the Bayes factors. The first set of parameters is taken as the reference model so its estimate is always 0.
The value of the loglikelihood at different samples and different parameters. This should be entered as a matrix where the rows are the values of the samples and the columns correspond to the parameters. The [i,j] element of the matrix is the value of the loglikelihood at the ith sample when all samples are put together evaluated at the jth parameter value.
A vector of length ncol(lglk) or a scalar corresponding to the sample sizes from each model. Must sum(N) == nrow(lglk). The first N[1] samples come from model corresponding to the first set of parameters, then (N[1]+1):N[2] are from the model corresponding to the second set of parameters, and so on.
Estimation is done by maximising the reverse logistic log likelihood.
Geyer, C. J. (1994). Estimating normalizing constants and reweighting mixtures in Markov chain Monte Carlo. Technical Report 568, School of Statistics, University of Minnesota.